Delphi Then

• In 2012, we formed Delphi to “develop the theory and practice of epidemic forecasting, and its role in decision-making”
• Participated in annual CDC flu forecasting challenges since 2013, earned top place in several
• Awarded CDC’s National Center of Excellence for flu forecasting in 2019
• Published open code and data, including numerous influenza surveillance streams

Delphi Now

• We’ve gone from 7 members to over 40, to support the US COVID-19 response
• Added many smart folks from CMU, but also Stanford, UC Davis, USC, Google
• Beyond (before?) forecasting: we’ve been focusing, now more than ever, on data
• This problem is so hard that our best weapon is a diversity of data sources

COVIDcast Indicators

COVIDcast Ecosystem

This Talk

We can’t cover all of this! We’ll focus on our API, and some basic demos with our survey data (please ask about medical claims data, or ask about forecasting or nowcasting, during the Q and A)

Outline:

1. COVIDcast API
2. Symptom survey
3. Forecasting demo

Reproducible talk: all code included

COVIDcast API

The COVIDcast API is based on HTTP GET queries and returns data in JSON form. The base URL is https://api.covidcast.cmu.edu/epidata/api.php?source=covidcast

R and Python Packages

We also provide R and Python packages for API access. Highlights:

• Easy API querying:
  • Same specification structure (many default parameters)
  • Full support for data revisions (as of, issues, lag)
• Basic signal processing:
  • Correlations sliced by location, or by time
  • Data wrangling: preparing signals for analysis
• Plotting functionality:
  • Choropleth and bubble maps
  • Time series plots

(Have an idea? File an issue or contribute a PR on our public GitHub repo)

Example: Deaths

How many COVID-19 deaths have been reported per day, in my state, since March 1?

library(covidcast)
start_day = "2020-03-01"
end_day = "2020-12-15"
deaths = covidcast_signal(data_source = "usa-facts", 
                          signal = "deaths_7dav_incidence_num", 
                          start_day = start_day, end_day = end_day,
                          geo_type = "state", geo_values = "pa")

plot(deaths, plot_type = "line", 
     title = "New COVID-19 deaths in PA (7-day average)") + 
  scale_x_date(date_breaks = "1 month", date_labels = "%b %Y") +
  theme(legend.position = "none")

Example: Hospitalizations

What percentage of daily hospital admissions are due to COVID-19 in PA, NY, TX?

hosp = covidcast_signal(data_source = "hospital-admissions", 
                        signal = "smoothed_adj_covid19_from_claims",
                        start_day = start_day, end_day = end_day,
                        geo_type = "state", geo_values = c("pa", "ny", "tx"))

plot(hosp, plot_type = "line", 
     title = "% of hospital admissions due to COVID-19") + 
  geom_dl(aes(y = value, color = geo_value, label = toupper(geo_value)), 
          method = "last.bumpup") +
  scale_x_date(date_breaks = "1 month", date_labels = "%b %Y") +
  theme(legend.position = "none")

Example: Cases

What does the current COVID-19 incident case rate look like, nationwide?

cases = covidcast_signal(data_source = "usa-facts", 
                         signal = "confirmed_7dav_incidence_prop",
                         start_day = end_day, end_day = end_day)

plot(cases, title = "New COVID-19 cases per 100,000 people")

Example: Cases (Cont.)

What does the current COVID-19 cumulative case rate look like, nationwide?

cases = covidcast_signal(data_source = "usa-facts", 
                         signal = "confirmed_cumulative_prop",
                         start_day = end_day, end_day = end_day)

plot(cases, title = "Cumulative COVID-19 cases per 100,000 people", 
     choro_params = list(legend_n = 6))

Example: Doctor’s Visits

How do some cities compare in terms of doctor’s visits due to COVID-like illness?

cities = c("Miami", "New York", "Pittsburgh", "San Antonio")
dv = covidcast_signal(data_source = "doctor-visits", 
                      signal = "smoothed_adj_cli", 
                      start_day = start_day, end_day = end_day,
                      geo_type = "msa", geo_values = name_to_cbsa(cities))

plot(dv, plot_type = "line", 
     title = "% of doctor's visits due to COVID-like illness") + 
  scale_x_date(date_breaks = "1 month", date_labels = "%b %Y")  +
  scale_color_hue(labels = cities)

Example: Symptoms

How do my county and my friend’s county compare in terms of COVID symptoms?

counties = c("Allegheny", "Fulton")
sympt = covidcast_signal(data_source = "fb-survey", 
                         signal = "smoothed_hh_cmnty_cli",
                         start_day = "2020-04-15", end_day = end_day,
                         geo_values = c(name_to_fips(counties[1]),
                                        name_to_fips(counties[2], 
                                                     state = "GA")))

plot(sympt, plot_type = "line", 
     title = "% of people who know somebody with COVID symptoms") + 
  scale_x_date(date_breaks = "1 month", date_labels = "%b %Y") +
  scale_color_hue(labels = counties)

Example: Mask Use

How do some states compare in terms of self-reported mask useage?

mask = covidcast_signal(data_source = "fb-survey", 
                        signal = "smoothed_wwearing_mask",
                        start_day = "2020-09-15", end_day = end_day,
                        geo_type = "state", 
                        geo_values = c("dc", "ma", "ny",
                                       "wy", "sd", "id"))

plot(mask, plot_type = "line", 
     title = "% of people who wear masks in public most/all the time") +
  geom_dl(aes(y = value, color = geo_value, label = toupper(geo_value)), 
          method = "last.bumpup") +
  scale_x_date(date_breaks = "1 month", date_labels = "%b %Y") +
  theme(legend.position = "none")

As Of, Issues, Lag

By default the API returns the most recent data for each time_value. We also provide access to all previous versions of the data, using the following optional parameters:

Data Revisions

Why would we need this? Because many data sources are subject to revisions:

• Case and death counts frequently corrected/adjusted by authorities
• Medical claims data can take weeks to be submitted and/or processed
• Testing/lab data can be backlogged for a variety of reasons, etc.

This presents a challenge to modelers: e.g., we have to learn how to forecast based on the data we’d have at the time, not updates that would arrive later

To accommodate, we log revisions even when the original data source does not!

Example: Backfill in Doctor’s Visits

The last two weeks of August in CA …

# Let's get the data that was available as of 09/22, for the end of August in CA
dv = covidcast_signal(data_source = "doctor-visits", 
                      signal = "smoothed_adj_cli",
                      start_day = "2020-08-15", end_day = "2020-08-31",
                      geo_type = "state", geo_values = "ca",
                      as_of = "2020-09-21")

# Plot the time series curve
xlim = c(as.Date("2020-08-15"), as.Date("2020-09-21"))
ylim = c(3.83, 5.92)
ggplot(dv, aes(x = time_value, y = value)) + 
  geom_line() +
  coord_cartesian(xlim = xlim, ylim = ylim) +
  geom_vline(aes(xintercept = as.Date("2020-09-21")), lty = 2) +
  labs(color = "as of", x = "Date", y = "% doctor's visits due to CLI in CA") +
  theme_bw() + theme(legend.position = "bottom")

Example: Backfill in Doctor’s Visits (Cont.)

The last two weeks of August in CA …

# Now loop over a bunhch of "as of" dates, fetch data from the API for each one
as_ofs = seq(as.Date("2020-09-01"), as.Date("2020-09-21"), by = "week")
dv_as_of = map_dfr(as_ofs, function(as_of) {
  covidcast_signal(data_source = "doctor-visits", signal = "smoothed_adj_cli",
                   start_day = "2020-08-15", end_day = "2020-08-31", 
                   geo_type = "state", geo_values = "ca", as_of = as_of)
})

# Plot the time series curve "as of" September 1
dv_as_of %>% 
  filter(issue == as.Date("2020-09-01")) %>% 
  ggplot(aes(x = time_value, y = value)) + 
  geom_line(aes(color = factor(issue))) + 
  coord_cartesian(xlim = xlim, ylim = ylim) +
  geom_vline(aes(color = factor(issue), xintercept = issue), lty = 2) +
  labs(color = "as of", x = "Date", y = "% doctor's visits due to CLI in CA") +
  geom_line(data = dv, aes(x = time_value, y = value)) +
  geom_vline(aes(xintercept = as.Date("2020-09-21")), lty = 2) +
  theme_bw() + theme(legend.position = "none")

Example: Backfill in Doctor’s Visits (Cont.)

The last two weeks of August in CA …

dv_as_of %>% 
  ggplot(aes(x = time_value, y = value)) + 
  geom_line(aes(color = factor(issue))) + 
  coord_cartesian(xlim = xlim, ylim = ylim) +
  geom_vline(aes(color = factor(issue), xintercept = issue), lty = 2) +
  labs(color = "as of", x = "Date", y = "% doctor's visits due to CLI in CA") +
  geom_line(data = dv, aes(x = time_value, y = value)) +
  geom_vline(aes(xintercept = as.Date("2020-09-21")), lty = 2) +
  theme_bw() + theme(legend.position = "none")

Massive Symptom Survey

Through recruitment partnership with Facebook, we survey about 50,000 people daily (and over 14 million since it began in April), in the United States about:

• COVID symptoms (common and rare)
• COVID testing
• Mental health
• Social contacts and behavior
• Demographics

A parallel, international effort by the University of Maryland reaches 100+ countries in 55 languages

Massive Symptom Survey (Cont.)

• The survey is implemented on Qualtrics, managed by CMU
• Facebook never receives any individual responses
• Designed to take about 10 minutes; about 35 questions
• Questions selected for relevance to forecasting and nowcasting
• But also for research in public health (understanding the pandemic burden)
• Respondents provide ZIP code, so data is available at county level

This is the largest non-Census research survey ever conducted (that we know of). Raw response data is freely available to researchers who sign a data use agreement

COVID-Like Illness

Using the survey data we generate daily, county-level estimates of:

• % CLI: the percentage of people with COVID-like illness
• % CLI-in-community: the percentage of people who know someone in their local community with COVID-like illness

(Note that COVID-like illness or CLI is defined as fever of at least 100 °F, along with cough, shortness of breath, or difficulty breathing. We also ask people to report on more rare symptoms)

Why % CLI-in-Community?

Why ask a proxy question (have people report on others)? Here’s Spearman correlations to COVID-19 case rates (sliced by time):

# Fetch Facebook % CLI signal, % CLI-in-community signal and confirmed case
# incidence proportions
start_day = "2020-04-15"
end_day = "2020-12-15"
sympt1 = covidcast_signal("fb-survey", "smoothed_cli", 
                          start_day, end_day)
sympt2 = covidcast_signal("fb-survey", "smoothed_hh_cmnty_cli", 
                          start_day, end_day)
cases = covidcast_signal("usa-facts", "confirmed_7dav_incidence_prop", 
                         start_day, end_day)

# Consider only counties with at least 500 cumulative cases so far
case_num = 500
geo_values = covidcast_signal("usa-facts", "confirmed_cumulative_num",
                              max(cases$time), max(cases$time)) %>%
  filter(value >= case_num) %>% pull(geo_value)
sympt1_act = sympt1 %>% filter(geo_value %in% geo_values)
sympt2_act = sympt2 %>% filter(geo_value %in% geo_values)
cases_act = cases %>% filter(geo_value %in% geo_values)

# Compute correlations, per time, over all counties
df_cor1 = covidcast_cor(sympt1_act, cases_act, by = "time_value", 
                        method = "spearman")
df_cor2 = covidcast_cor(sympt2_act, cases_act, by = "time_value", 
                        method = "spearman")

# Stack rowwise into one data frame
df_cor = rbind(df_cor1, df_cor2)
df_cor$signal = c(rep("% CLI", nrow(df_cor1)), 
                  rep("% CLI-in-community", nrow(df_cor2)))

# Then plot correlations over time 
ggplot(df_cor, aes(x = time_value, y = value)) + 
  geom_line(aes(color = signal)) +
  labs(title = "Correlation between survey signals and case rates (by time)",
       subtitle = sprintf("Over all counties with at least %i cumulative cases",
                          case_num), x = "Date", y = "Correlation") +
    theme_bw() + theme(legend.pos = "bottom", legend.title = element_blank())

Beyond Symptom Data

Reminder: survey data extends far beyond symptoms. For example:

• Test positivity rate, broken down by test reason
• % of people tested in last 14 days, broken down by age, occupation
• % of people ever tested
• Types of activities people do outside their homes
• Mask wearing, reporting on self/others in community
• Anxiety, depression, isolation

Example: Mask Use in Others

Interesting discrepancy in reported use by self versus others … but not unexpected?

start_day = "2020-12-01"
end_day = "2020-12-01"
signals = covidcast_signals(data_source = "fb-survey",
                            signal = c("smoothed_wwearing_mask",
                                       "smoothed_wothers_masked"),
                            start_day, end_day, geo_type = "state")

joined = inner_join(signals[[1]], signals[[2]], by = "geo_value",
                    suffix = c(".self", ".others")) 

perc = seq(min(signals[[1]]$value), max(signals[[1]]$value), length = 100)
binom = data.frame(x = perc, y = 100 * pbinom(16, 20, perc/100, lower = FALSE))

ggplot(joined, aes(x = value.self, y = value.others)) +
  geom_text(aes(label = toupper(geo_value))) +
  geom_line(data = binom, mapping = aes(x = x, y = y)) +
  labs(x = "% who report wearing masks most/all the time",
       y = "% who report most/all others wear masks",
       title = "Mask use reported in symptom survey") +
  theme_bw()

An Early Indicator?

As motivation, let’s take a look at case counts in Miami-Dade, from June 1 to July 15, and compare it to the % CLI-in-community indicator based on our survey:

# Fetch Facebook % CLI-in-community signal and confirmed case incidence numbers
# from June 1 to July 15
start_day = "2020-06-01"
end_day = "2020-07-15"
sympt = covidcast_signal("fb-survey", "smoothed_hh_cmnty_cli", 
                         start_day, end_day)
cases = covidcast_signal("usa-facts", "confirmed_7dav_incidence_num",
                         start_day, end_day)

# Function to transform from one range to another
trans = function(x, from_range, to_range) {
  (x - from_range[1]) / (from_range[2] - from_range[1]) *
    (to_range[2] - to_range[1]) + to_range[1]
}

# Function to produce a plot comparing the signals for one county
ggplot_colors = c("#FC4E07", "#00AFBB", "#E7B800")
plot_one = function(geo_value, df1, df2, lab1, lab2, title = NULL, 
                    xlab = NULL, ylab1 = NULL, ylab2 = NULL) {
  # Filter down the signal data frames
  given_geo_value = geo_value
  df1 = df1 %>% filter(geo_value == given_geo_value)
  df2 = df2 %>% filter(geo_value == given_geo_value)
  
  # Compute ranges of the two signals
  range1 = df2 %>% select("value") %>% range(na.rm = TRUE)
  range2 = df1 %>% select("value") %>% range(na.rm = TRUE)
  
  # Convenience functions for our two signal ranges
  trans12 = function(x) trans(x, range1, range2)
  trans21 = function(x) trans(x, range2, range1)
  
  # Find state name, find abbreviation, then set title
  state_name = fips_to_name(paste0(substr(geo_value, 1, 2), "000"))
  state_abbr = name_to_abbr(state_name)
  title = paste0(fips_to_name(geo_value), ", ", state_abbr)
  
  # Transform the combined signal to the incidence range, then stack
  # these rowwise into one data frame
  df = select(rbind(df1 %>% mutate_at("value", trans21),
                    df2), c("time_value", "value"))
  df$signal = c(rep(lab1, nrow(df1)), rep(lab2, nrow(df2)))
  
  # Finally, plot both signals
  return(ggplot(df, aes(x = time_value, y = value)) +
           geom_line(aes(color = signal)) +
           scale_color_manual(values = ggplot_colors[1:2]) +
           scale_y_continuous(name = ylab1, limits = range1,
                              sec.axis = sec_axis(trans = trans12,
                                                  name = ylab2)) +
           labs(title = title, x = xlab) + theme_bw() +
           theme(legend.pos = "bottom", legend.title = element_blank()))
}

# Produce a plot for Miami-Dade, and add vertical lines
plot_one(name_to_fips("Miami-Dade"), df1 = sympt, df2 = cases, 
         lab1 = "% CLI-in-community", lab2 = "New COVID-19 cases", 
         xlab = "Date", ylab1 = "New COVID-19 cases",
         ylab2 = "% of people who know someone with CLI") +
  geom_vline(xintercept = as.numeric(as.Date("2020-06-19")),
             linetype = 2, size = 1, color = ggplot_colors[1]) +
  geom_vline(xintercept = as.numeric(as.Date("2020-06-25")),
             linetype = 2, size = 1, color = ggplot_colors[2])

An Early Indicator? (Cont.)

Let’s look again, now at Allegheny County, right around the third wave:

start_day = "2020-10-15"
end_day = "2020-12-01"
sympt = covidcast_signal("fb-survey", "smoothed_hh_cmnty_cli", 
                         start_day, end_day)
cases = covidcast_signal("usa-facts", "confirmed_7dav_incidence_num",
                         start_day, end_day)

# Produce a plot for Allegheny County, and add vertical lines
plot_one(name_to_fips("Allegheny"), df1 = sympt, df2 = cases,
         lab1 = "% CLI-in-community", lab2 = "New COVID-19 cases", 
         xlab = "Date", ylab1 = "New COVID-19 cases",
         ylab2 = "% of people who know someone with CLI") +
  geom_vline(xintercept = as.numeric(as.Date("2020-10-30")),
             linetype = 2, size = 1, color = ggplot_colors[1]) +
  geom_vline(xintercept = as.numeric(as.Date("2020-11-06")),
             linetype = 2, size = 1, color = ggplot_colors[2])

An Early Indicator? (Cont.)

Once more, still Allegheny County, now % of doctor visits with CLI:

dv = covidcast_signal("chng", "smoothed_adj_outpatient_cli", 
                      start_day, end_day)

# Produce a plot for Allegheny County, and add vertical lines
plot_one(name_to_fips("Allegheny"), df1 = dv, df2 = cases,
         lab1 = "% CLI-in-doctor-visits", lab2 = "New COVID-19 cases", 
         xlab = "Date", ylab1 = "New COVID-19 cases",
         ylab2 = "% of people who know someone with CLI") +
  geom_vline(xintercept = as.numeric(as.Date("2020-11-02")),
             linetype = 2, size = 1, color = ggplot_colors[1]) +
  geom_vline(xintercept = as.numeric(as.Date("2020-11-06")),
             linetype = 2, size = 1, color = ggplot_colors[2])

Simple Forecasting Demo

Notational setup: for location (county) \(\ell\) and time (day) \(t\), let

• \(Y_{\ell,t}\) denote the COVID-19 case incidence rate,
• \(X_{\ell,t}\) the % CLI-in-community or % CLI-in-doctor-visits signal.

To predict case rates \(d\) days ahead, consider two simple models: \[ \begin{align*} & h(Y_{\ell,t+d}) \approx \alpha + \sum_{j=0}^2 \beta_j h(Y_{\ell,t-7j}) \quad \text{(Cases)} \\ & h(Y_{\ell,t+d}) \approx \alpha + \sum_{j=0}^2 \beta_j h(Y_{\ell,t-7j}) + \sum_{j=0}^2 \gamma_j h(X_{\ell,t-7j}) \quad \text{(Cases + signal)} \\ \end{align*} \]

For each forecast date, we train models on the most recent 14 days worth of data (setup adapted from this blog post)

Simple Forecasting Demo (Cont.)

Results from forecasts made over early May to October 15:

# This RData file was too big to upload to GitHub, but it can be produced by
# running the forecasting demo in "demo.R"
load("demo.rda")

# Compute and plot median errors as function of number of days ahead
err_by_lead = res %>%
  drop_na() %>%                                       # Restrict to common time
  mutate(err1 = err1 / err0, err2 = err2 / err0,
         err3 = err3 / err0, err4 = err4 / err0) %>%  # Compute relative error
                                                      # to strawman model
  ungroup() %>%
  select(-c(err0)) %>%
  pivot_longer(names_to = "model", values_to = "err",
               cols = -c(geo_value, time_value, lead)) %>%
  mutate(model = factor(model, labels = c("Cases", 
                                          "Cases + CLI-in-community",
                                          "Cases + CLI-in-doctor-visits",
                                          "Cases + both signals")),
         time_period = ifelse(time_value <= "2020-10-15", 1, 2)) %>%
  group_by(model, time_period, lead) %>%
  summarize(err = median(err)) %>% 
  ungroup()

ggplot(err_by_lead %>% filter(time_period == 1), aes(x = lead, y = err)) + 
  geom_line(aes(color = model)) + 
  geom_point(aes(color = model)) + 
  geom_hline(yintercept = err_by_lead %>% 
               filter(time_period == 1, lead %in% 7, model == "Cases") %>% 
               pull(err), linetype = 2, color = "gray") +
  scale_color_manual(values = c("black", ggplot_colors)) +
  labs(title = "Forecasting errors by number of days ahead",
       subtitle = "From early May through October 15",
       x = "Number of days ahead", y = "Median scaled error") +
  theme_bw() + theme(legend.pos = "bottom", legend.title = element_blank())

Simple Forecasting Demo (Cont.)

Results from forecasts made over October 15 to current day:

ggplot(err_by_lead %>% filter(time_period == 2), aes(x = lead, y = err)) + 
  geom_line(aes(color = model)) + 
  geom_point(aes(color = model)) + 
  scale_color_manual(values = c("black", ggplot_colors)) +
  labs(title = "Forecasting errors by number of days ahead",
       subtitle = "From early October 15 through current day",
       x = "Number of days ahead", y = "Median scaled error") +
  theme_bw() + theme(legend.pos = "bottom", legend.title = element_blank())

Wrapping Up

Delphi’s COVIDcast ecosystem has many parts:

1. Unique relationships with partners in tech and healthcare granting us access to data on pandemic activity
2. Code and infrastructure to build COVID-19 indicators, continuously-updated and geographically-comprehensive
3. A historical database of all indicators, including revision tracking
4. A public API (and R and Python packages) serving new indicators daily
5. Interactive maps and graphics to display our indicators
6. Nowcasting and forecasting work building on the indicators

In this pandemic, it’ll take an entire community to find answers to all the important questions. Please join ours!

Thanks

• The whole Delphi team, and various CMU units
• Google, Facebook, Amazon Web Services
• Change Healthcare, Quidel, SafeGraph, Qualtrics
• Centers for Disease Control and Prevention

Go to: https://covidcast.cmu.edu … you’ll find everything linked from there!

Helpful Links

• API documentation
• Client packages: R and Python
• Signal documentation
• Survey documentation
• Interactive graphics
• Delphi’s blog
• Delphi’s GitHub org

Example API Query

Estimated % COVID-like illness on April 6, 2020 from the Facebook survey, in Allegheny County: https://api.covidcast.cmu.edu/epidata/api.php?source=covidcast&data_source=fb-survey&signal=raw_cli&time_type=day&geo_type=county&time_values=20200406&geo_value=42003

library(jsonlite)
res = readLines("https://api.covidcast.cmu.edu/epidata/api.php?source=covidcast&data_source=fb-survey&signal=raw_cli&time_type=day&geo_type=county&time_values=20200406&geo_value=42003")
prettify(res)

## {
##     "result": 1,
##     "epidata": [
##         {
##             "geo_value": "42003",
##             "signal": "raw_cli",
##             "time_value": 20200406,
##             "direction": null,
##             "issue": 20200903,
##             "lag": 150,
##             "value": 0.7614984,
##             "stderr": 0.3826746,
##             "sample_size": 434.8891
##         }
##     ],
##     "message": "success"
## }
## 

Current Metadata

meta = covidcast_meta()
summary(meta)

## A `covidcast_meta` data frame with 613 rows and 15 columns.
## 
## Number of data sources : 14
## Number of signals      : 155
## 
## Summary:
## 
##  data_source           signal                                           county
##  chng                  smoothed_adj_outpatient_cli                      *     
##  chng                  smoothed_adj_outpatient_covid                    *     
##  chng                  smoothed_outpatient_cli                          *     
##  chng                  smoothed_outpatient_covid                        *     
##  doctor-visits         smoothed_adj_cli                                 *     
##  doctor-visits         smoothed_cli                                     *     
##  fb-survey             raw_cli                                          *     
##  fb-survey             raw_hh_cmnty_cli                                 *     
##  fb-survey             raw_ili                                          *     
##  fb-survey             raw_nohh_cmnty_cli                               *     
##  fb-survey             raw_wcli                                         *     
##  fb-survey             raw_whh_cmnty_cli                                *     
##  fb-survey             raw_wili                                         *     
##  fb-survey             raw_wnohh_cmnty_cli                              *     
##  fb-survey             smoothed_anxious_5d                              *     
##  fb-survey             smoothed_cli                                     *     
##  fb-survey             smoothed_depressed_5d                            *     
##  fb-survey             smoothed_felt_isolated_5d                        *     
##  fb-survey             smoothed_hh_cmnty_cli                            *     
##  fb-survey             smoothed_ili                                     *     
##  fb-survey             smoothed_large_event_1d                          *     
##  fb-survey             smoothed_nohh_cmnty_cli                          *     
##  fb-survey             smoothed_others_masked                           *     
##  fb-survey             smoothed_public_transit_1d                       *     
##  fb-survey             smoothed_restaurant_1d                           *     
##  fb-survey             smoothed_shop_1d                                 *     
##  fb-survey             smoothed_spent_time_1d                           *     
##  fb-survey             smoothed_tested_14d                              *     
##  fb-survey             smoothed_tested_positive_14d                     *     
##  fb-survey             smoothed_travel_outside_state_5d                 *     
##  fb-survey             smoothed_wanted_test_14d                         *     
##  fb-survey             smoothed_wanxious_5d                             *     
##  fb-survey             smoothed_wcli                                    *     
##  fb-survey             smoothed_wdepressed_5d                           *     
##  fb-survey             smoothed_wearing_mask                            *     
##  fb-survey             smoothed_wfelt_isolated_5d                       *     
##  fb-survey             smoothed_whh_cmnty_cli                           *     
##  fb-survey             smoothed_wili                                    *     
##  fb-survey             smoothed_wlarge_event_1d                         *     
##  fb-survey             smoothed_wnohh_cmnty_cli                         *     
##  fb-survey             smoothed_work_outside_home_1d                    *     
##  fb-survey             smoothed_worried_become_ill                      *     
##  fb-survey             smoothed_worried_finances                        *     
##  fb-survey             smoothed_wothers_masked                          *     
##  fb-survey             smoothed_wpublic_transit_1d                      *     
##  fb-survey             smoothed_wrestaurant_1d                          *     
##  fb-survey             smoothed_wshop_1d                                *     
##  fb-survey             smoothed_wspent_time_1d                          *     
##  fb-survey             smoothed_wtested_14d                             *     
##  fb-survey             smoothed_wtested_positive_14d                    *     
##  fb-survey             smoothed_wtravel_outside_state_5d                *     
##  fb-survey             smoothed_wwanted_test_14d                        *     
##  fb-survey             smoothed_wwearing_mask                           *     
##  fb-survey             smoothed_wwork_outside_home_1d                   *     
##  fb-survey             smoothed_wworried_become_ill                     *     
##  fb-survey             smoothed_wworried_finances                       *     
##  ght                   raw_search                                             
##  ght                   smoothed_search                                        
##  google-survey         raw_cli                                          *     
##  google-survey         smoothed_cli                                     *     
##  google-symptoms       ageusia_raw_search                               *     
##  google-symptoms       ageusia_smoothed_search                          *     
##  google-symptoms       anosmia_raw_search                               *     
##  google-symptoms       anosmia_smoothed_search                          *     
##  google-symptoms       sum_anosmia_ageusia_raw_search                   *     
##  google-symptoms       sum_anosmia_ageusia_smoothed_search              *     
##  hospital-admissions   smoothed_adj_covid19                             *     
##  hospital-admissions   smoothed_adj_covid19_from_claims                 *     
##  hospital-admissions   smoothed_covid19                                 *     
##  hospital-admissions   smoothed_covid19_from_claims                     *     
##  indicator-combination confirmed_7dav_cumulative_num                    *     
##  indicator-combination confirmed_7dav_cumulative_prop                   *     
##  indicator-combination confirmed_7dav_incidence_num                     *     
##  indicator-combination confirmed_7dav_incidence_prop                    *     
##  indicator-combination confirmed_cumulative_num                         *     
##  indicator-combination confirmed_cumulative_prop                        *     
##  indicator-combination confirmed_incidence_num                          *     
##  indicator-combination confirmed_incidence_prop                         *     
##  indicator-combination deaths_7dav_cumulative_num                       *     
##  indicator-combination deaths_7dav_cumulative_prop                      *     
##  indicator-combination deaths_7dav_incidence_num                        *     
##  indicator-combination deaths_7dav_incidence_prop                       *     
##  indicator-combination deaths_cumulative_num                            *     
##  indicator-combination deaths_cumulative_prop                           *     
##  indicator-combination deaths_incidence_num                             *     
##  indicator-combination deaths_incidence_prop                            *     
##  indicator-combination nmf_day_doc_fbc_fbs_ght                          *     
##  indicator-combination nmf_day_doc_fbs_ght                              *     
##  jhu-csse              confirmed_7dav_cumulative_num                    *     
##  jhu-csse              confirmed_7dav_cumulative_prop                   *     
##  jhu-csse              confirmed_7dav_incidence_num                     *     
##  jhu-csse              confirmed_7dav_incidence_prop                    *     
##  jhu-csse              confirmed_cumulative_num                         *     
##  jhu-csse              confirmed_cumulative_prop                        *     
##  jhu-csse              confirmed_incidence_num                          *     
##  jhu-csse              confirmed_incidence_prop                         *     
##  jhu-csse              deaths_7dav_cumulative_num                       *     
##  jhu-csse              deaths_7dav_cumulative_prop                      *     
##  jhu-csse              deaths_7dav_incidence_num                        *     
##  jhu-csse              deaths_7dav_incidence_prop                       *     
##  jhu-csse              deaths_cumulative_num                            *     
##  jhu-csse              deaths_cumulative_prop                           *     
##  jhu-csse              deaths_incidence_num                             *     
##  jhu-csse              deaths_incidence_prop                            *     
##  nchs-mortality        deaths_allcause_incidence_num                          
##  nchs-mortality        deaths_allcause_incidence_prop                         
##  nchs-mortality        deaths_covid_and_pneumonia_notflu_incidence_num        
##  nchs-mortality        deaths_covid_and_pneumonia_notflu_incidence_prop       
##  nchs-mortality        deaths_covid_incidence_num                             
##  nchs-mortality        deaths_covid_incidence_prop                            
##  nchs-mortality        deaths_flu_incidence_num                               
##  nchs-mortality        deaths_flu_incidence_prop                              
##  nchs-mortality        deaths_percent_of_expected                             
##  nchs-mortality        deaths_pneumonia_notflu_incidence_num                  
##  nchs-mortality        deaths_pneumonia_notflu_incidence_prop                 
##  nchs-mortality        deaths_pneumonia_or_flu_or_covid_incidence_num         
##  nchs-mortality        deaths_pneumonia_or_flu_or_covid_incidence_prop        
##  quidel                covid_ag_raw_pct_positive                        *     
##  quidel                covid_ag_smoothed_pct_positive                   *     
##  quidel                raw_pct_negative                                       
##  quidel                raw_tests_per_device                                   
##  quidel                smoothed_pct_negative                                  
##  quidel                smoothed_tests_per_device                              
##  safegraph             bars_visit_num                                   *     
##  safegraph             bars_visit_prop                                  *     
##  safegraph             completely_home_prop                             *     
##  safegraph             completely_home_prop_7dav                        *     
##  safegraph             full_time_work_prop                              *     
##  safegraph             full_time_work_prop_7dav                         *     
##  safegraph             median_home_dwell_time                           *     
##  safegraph             median_home_dwell_time_7dav                      *     
##  safegraph             part_time_work_prop                              *     
##  safegraph             part_time_work_prop_7dav                         *     
##  safegraph             restaurants_visit_num                            *     
##  safegraph             restaurants_visit_prop                           *     
##  usa-facts             confirmed_7dav_cumulative_num                    *     
##  usa-facts             confirmed_7dav_cumulative_prop                   *     
##  usa-facts             confirmed_7dav_incidence_num                     *     
##  usa-facts             confirmed_7dav_incidence_prop                    *     
##  usa-facts             confirmed_cumulative_num                         *     
##  usa-facts             confirmed_cumulative_prop                        *     
##  usa-facts             confirmed_incidence_num                          *     
##  usa-facts             confirmed_incidence_prop                         *     
##  usa-facts             deaths_7dav_cumulative_num                       *     
##  usa-facts             deaths_7dav_cumulative_prop                      *     
##  usa-facts             deaths_7dav_incidence_num                        *     
##  usa-facts             deaths_7dav_incidence_prop                       *     
##  usa-facts             deaths_cumulative_num                            *     
##  usa-facts             deaths_cumulative_prop                           *     
##  usa-facts             deaths_incidence_num                             *     
##  usa-facts             deaths_incidence_prop                            *     
##  youtube-survey        raw_cli                                                
##  youtube-survey        raw_ili                                                
##  youtube-survey        smoothed_cli                                           
##  youtube-survey        smoothed_ili                                           
##  msa hrr state
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *       *    
##  *       *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##          *    
##  *   *   *    
##  *   *   *    
##  *       *    
##  *       *    
##  *       *    
##  *       *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##  *   *   *    
##          *    
##          *    
##          *    
##          *

Access to Survey Microdata

Want to study a problem that can be answered with over 14 million US survey responses since April? Possible topics:

• Symptoms reported by people testing positive, stratified by chronic conditions, age, etc.
• Test rates and availability by employment and occupation
• Mental health impacts of interventions
• Disparate impacts on minorities and disadvantaged groups
• … anything else you can think of

Raw response data is freely available to researchers who sign a data use agreement to protect confidentiality of responses

We’re building a network of academic and non-profit researchers to learn from the survey. Join us!